Few words stir up more controversy in an industrial inspection lab than “CMM TRAINING.” Everyone has an opinion on it but few have any facts without paying a steep price. The price may be in costly training that doesn’t produce the level of competency that your facility expects. Sometimes that cost increases by the denial that your training (and your facility’s follow-up) stirred up more questions (and more controversy) than it did provide answers or solutions. The bottom line is that your facility and it’s managers must admit they don’t know (what they don’t know) about what CMM training should consists of and are better off stepping out of the way and letting those who do know, plan the training and develop the training plan. That being said, if you don’t know what you don’t know… read and learn.

Even though the quality of probe calibration and compensation can determine how well a Coordinate Measuring Machine does or does not measure on every given day, the probe compensation process is often overlooked as automatic. Effective training will discuss the phenomena at work as your CMM compensates the radius of the specific (normally ruby) stylus being used and simple mathematical tests that can be used to control and confirm this process. This will become the cornerstone of your CMM measurements. This process is normally very precise and repeatable but leaving little to chance, very simple statistical control methods should be used to confirm the probe compensation before part measurements are taken. These statistical tools can be explained using nothing more than very basic case studies and a $10 calculator.

There are different coordinate systems used during CMM measuring. This will begin with the explanation of the Machine coordinate system (MCS) along with the Part coordinate system (PCS). Within the Part coordinate system several expressions that are commonly needed and used should be discussed and explained. Cartesian coordinates, Polar coordinates, (and perhaps Spherical coordinates) and how they relate and translate should be explained using High School level basic Mathematics. Too often CMM operators become prisoners within the coordinate system most frequently used with the common parts they measure. This approach limits their understanding of a CMM and prevents them from learning techniques that may, and most likely will, be needed in the not too distant future.

Recreating the specific part alignment to the Primary datum, Secondary datum, and applied Origin (as called out on the engineering drawing of a part) cannot be over emphasized. This basic alignment, or Part Coordinate System (PCS), can be explained using the 3-2-1 concept. The Baseplane is represented as the 3 points (minimum) needed to measure, construct, or create a plane. This alignment is often called “leveling”. The secondary alignment (often called “clocking”) is represented by the 2 points (minimum) needed to measure or create a line feature, and the number 1 signifies the point of origin (or Origin point) of the Part Coordinate System. When this system is aligned correctly the X, Y, and Z origin of the Part Coordinate System becomes the center of the universe when viewed by the CMM software. This process should be demonstrated using a series of graphics and data examples from CMM software packages that are visual based and CMM packages that are based more on using the data of the measured features to confirm the alignment.

Most CMM software offers specific macro systems programmed into the software to create alignment of the Part Coordinate System using some familiar geometric coordinate system alignments. Some of the common macros that might be explained are:

Plane, Line, Line

Plane, Circle, Line

Plane, Circle, Circle

Cylinder, Plane, Circle

Plane, Plane, Plane

Cylinder, Plane, Line

Cylinder, Plane, Midpoint

These macros can save a CMM operator or programmer much time in the short term but also may not create a disciplined system of the CMM operator being able to visualize the Part drawing (and CMM set- up) correlation in the long term. The pros and cons of relying too heavily on macros should be discussed and explained so the operator(s) can make the most beneficial macro choices at their facility in both the short and long term.

Not long after the alignment of Part Coordinate System is clearly understood, a datum offset alignment or part rotation may be necessary on a given part. Some software allows the CMM operator to establish multiple Part Coordinate Systems that can be saved within the software before and after offsets or rotations. These systems can be re-called later to cross reference any needed calculations to different alignments and Datums specified on the engineering drawing of the part. This process of Part Alignment, saving the PCS, offsetting or rotating the PCS, saving the new PCS, and recalling each PCS when needed should be explained to introduce the reader to the “flow” of the different coordinate systems with their part drawings and CMM measurements.

A dilemma as old as CMM itself is “how many points to probe?” The industrial market had demanded that CMM software is able to calculate features using the minimum number of points needed to create a geometric feature. 3 points to make a plane or a circle, 2 points to measure a line, 6 points to create a cylinder, etc. A CMM operator with any practical experience rarely, if ever, recommends this methodology of using the minimum number of needed points. CMMs are creating or calculating all of the geometric data of a feature from the points that are probed or constructed. If the only 3 points are used for every circle it is impossible to account for any “real world” roundness variation in the form of the circle. When measuring in microns and millionths, as today’s tolerances require, - nothing is round, flat, or parallel. A sufficient amount of points need to be probed or constructed to account for this variation within the form and location of part features. The sufficient amount of points may change for different features, or for similar features that require different specifications, or for similar features with similar specifications but different nominal values. In a perfect world where time would never be a factor, we could just probe 1000 points for each feature to be confident the data was reliable. In the real world in becomes a balancing act of probing enough points to assure accurate data yet not “over measuring” which can lengthen the time required for part measuring, which then increases inspection costs. This section should introduce the operator to some very simple, proactive, studies in order to determine the optimum balance of how many points to probe when measuring nearly any feature, on any part, of any specific tolerance, of various nominal values, using the CMM.

Not long after a CMM operator or programmer begins measuring parts after re-creating the Part Coordinate Systems as described on the part drawing, Geometric Dimensioning and Tolerancing questions will soon surface concerning the reliability of the measurement data. CMMs measure, calculate, and tolerance features just as defined by the ASME Y14.5-1994 (The standard for Geometric Dimensioning and Tolerancing). In many cases, an alternate measurement technique may have been used in the past at the operator’s facility that produces more favorable measurement results. When these 2 approaches collide, the method that attains the out of tolerance readings will always be refuted. During these situations it becomes very important for a CMM operator to reference the G,D,&T standard along with the CMM measurement data and PCS alignment to “prove” the readings are reliable and the feature tolerances are attained as defined by the G,D,&T standard. This section should explain misconceptions about some common tolerances as well as the proper method of calculation for these often misunderstood requirements. Generic CMM “drills” using common artifact standards such as Gage blocks, Ring gages, etc. can be explained to allow the operator to set up sample measurements as defined by the G,D,&T standard to test and confirm how their CMM calculates the features on artifacts that represent near perfect sample parts. While the test measurements are being performed and the CMM is being confirmed the operator can also be learning the G,D,&T/ CMM correlation. These measurements can be documented for future use or demonstrated to satisfy and silence any CMM critics.

A good CMM trainer will spend some time discussing the “pure” geometry of what the CMM is “doing”. I was working with a CAD programmer once who was attempting to program an EDM machine to cut a symmetrical gear shaped part. The drawing listed several dimensional and profile specifications. The customer had revised some of these specifications in order to modify the gear. The EDM programmer realized the revised specifications conflicted one another to the point that Angle A could not be in tolerance without making Distance B fall outside this listed specification. The programmer looked at the drawing and said to me; “we can’t get there from here.” Sometimes when a CMM operator needs to re-create the PCS of the drawing and measure the characteristics of the part a conflict will occur. This can often present itself because the part is drawn “floating in space” in a free-state condition and when measured it stays in a free state but must be fixtured in order to be measured. When these situations occur constructions can be made from intersecting, projected, or translated (offset) features. This section can explain this process using some of the common and not so common feature constructions using geometric images and sample data from CMM measurements. These examples will demonstrate how we can use creative, coordinate engineering to get from here to there.

In many CMMs, I-J-K and L-M-N: Direction Vectors will be used as the compass that points your CMM in the right direction. These vectors are derived from 1 angle and 1 Trig function. Although used extensively with DCC (Direct Computer Controlled) CMMs, Direction vectors are very helpful to understand when using a manual CMM as well. This section can explain the concept of vectors as the “direction” a feature is oriented when compared to the X,Y,and Z axes. This often intimidating concept of vector calculation can be explained using the “90 minus-rule” which allows the operator to calculate the vector direction of any feature by knowing only 1 angle at which to feature compares to an axis. The cosine of the known angle is used to calculate X axis vector (I or L) and the 90 minus rule is used to determine the 2nd angle and its vector (J or M). It sound complicated, but if it was – yours truly would have never been able to learn it or teach it. If explained by the right trainer, it becomes beads on a string…

Most CMM software on the market today allows the operator to select different parameters in which data will be calculated and displayed. A good trainer will explain some of the common CMM environment options and how to determine (by your part characteristics) which environment choices are the best choices in certain situations. Some of these concepts that will be discuss are: Coordinate Systems review (Polar, Cartesian, Spherical), Angle calculation (Decimal Degrees and DMS), Angle description (0° to 360° or 0° to 180°/-180°), number of decimal places displayed, Algorithms (LSQ, Min Outer, Max Inner), etc. Examples of when each option could be the best choice will be discussed and explained.

Last, but not at all least, your training should cover some time to put it all together. If you operators are motivated, and your trainer is effective, things will start “clicking” about half way through day #2. Then the operators can get in front of the CMM and walk through some of items covered in the training in a way that fits within the spectrum of what they are going to be measuring. I would also make sure your trainer is willing to offer follow up support via phone or e-mail as the operators begin their learning curve. The relationship with your facility and your CMM trainer could be long term if you are careful not to abuse it. Most good CMM trainers look forward to hearing from their clients and working with them to solve their roadblocks. You can bet while the trainer was learning, he/she had a few people on “speed-dial” whom they called when things were tough. It’s one of the golden rules, you’ll get out of it exaclty what you put into it. So get up early, study hard, and never give up… there is ALWAYS another move.

Richard Clark is a Metrologist, CMM programmer, and CMM trainer from Indiana. Feel free to e-mail him at cmm_rockstar@yahoo.com